{
  "id": 308419,
  "title": "How to Submit",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/308419",
  "author_name": "",
  "post_date": "2022-02-18T15:00:56.848178100Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Apologies for the beginner question, I'm aware that the competition is over at this point but am just trying to learn for future reference. </p>\n<p>I have model that can be fed an image and output a prediction in the format asked, which is:<br>\n \"Confidence X Y Width Height\", but how does this translate to the submission note book that was given? The evaluation page gives a format that I'm only able to run in kaggle notebooks. Am I supposed to submit a zipped file that can intake files and output the predictions? Or do I write something in Kaggle notebooks with the uploaded model?</p>\n<p>Thanks</p>",
  "messages": [
    {
      "id": "1696055",
      "postDate": "02/18/2022 15:00:56",
      "content": "<p>Apologies for the beginner question, I'm aware that the competition is over at this point but am just trying to learn for future reference. </p>\n<p>I have model that can be fed an image and output a prediction in the format asked, which is:<br>\n \"Confidence X Y Width Height\", but how does this translate to the submission note book that was given? The evaluation page gives a format that I'm only able to run in kaggle notebooks. Am I supposed to submit a zipped file that can intake files and output the predictions? Or do I write something in Kaggle notebooks with the uploaded model?</p>\n<p>Thanks</p>",
      "rawMarkdown": "Apologies for the beginner question, I'm aware that the competition is over at this point but am just trying to learn for future reference. \n\nI have model that can be fed an image and output a prediction in the format asked, which is:\n \"Confidence X Y Width Height\", but how does this translate to the submission note book that was given? The evaluation page gives a format that I'm only able to run in kaggle notebooks. Am I supposed to submit a zipped file that can intake files and output the predictions? Or do I write something in Kaggle notebooks with the uploaded model?\n\nThanks",
      "votes": null
    },
    {
      "id": "1697517",
      "postDate": "02/19/2022 17:08:44",
      "content": "<p><a href=\"https://www.kaggle.com/jackhasselbring\" target=\"_blank\">@jackhasselbring</a> . You should upload the model you have trained as a dataset, and then use it to make inference on the test set directly in the notebook using the provided API. There are inference model examples in the notebook section. Good luck ! </p>",
      "rawMarkdown": "jackhasselbring . You should upload the model you have trained as a dataset, and then use it to make inference on the test set directly in the notebook using the provided API. There are inference model examples in the notebook section. Good luck !",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1697517,
      "author_name": "vladvdv",
      "author_url": "",
      "post_date": "02/19/2022 17:08:44",
      "content": "<p><a href=\"https://www.kaggle.com/jackhasselbring\" target=\"_blank\">@jackhasselbring</a> . You should upload the model you have trained as a dataset, and then use it to make inference on the test set directly in the notebook using the provided API. There are inference model examples in the notebook section. Good luck ! </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1696055": "Apologies for the beginner question, I'm aware that the competition is over at this point but am just trying to learn for future reference. \n\nI have model that can be fed an image and output a prediction in the format asked, which is:\n \"Confidence X Y Width Height\", but how does this translate to the submission note book that was given? The evaluation page gives a format that I'm only able to run in kaggle notebooks. Am I supposed to submit a zipped file that can intake files and output the predictions? Or do I write something in Kaggle notebooks with the uploaded model?\n\nThanks",
    "1697517": "jackhasselbring . You should upload the model you have trained as a dataset, and then use it to make inference on the test set directly in the notebook using the provided API. There are inference model examples in the notebook section. Good luck !"
  },
  "source": "meta"
}